Which model of MC method is implemented in AvalancheMC?

Garfield++
1

Hi I am studying about Avalanche method in garfield++

I wanna ask question about MC method implemented in garfield++

Normally MC method describing avalanche method is using band calculation but when i give a look at AvalancheMC code, it seems to use local electric field model and give randomness.

Can u let me know which kind of MC model is implemented in garfield++? It will be really helpful if you can give me name of MC model and reference

Always thanks for your help

3

I found reference but process describing avalanche in AvalanchMC is different comparing to well known monte carlo method.

Can i ask reason why garfield use different avalanche mechanism?

4

Hi,
thanks for the ping and sorry for the late reply!
You are right, the method implemented in AvalancheMC is not a “microscopic” Monte Carlo simulation. It uses macroscopic transport parameters like drift velocity, diffusion coefficients and Townsend coefficient to simulate drift lines.
If I understood your message correctly, a technique similar to what you have in mind is implemented in AvalancheMicroscopic but at the moment it only works for gases. I did try to extend it to silicon a few years ago but the code is not properly tested/debugged so it’s not really useable at the moment. It would be great to get it to work but that would require a significant effort.

5

Thanks for your kind answer! Actually I saw previous cern’s report about silicon simulation and their exist data about describing LGAD simulation using garfield++. So we can conclude that report is described with macroscopic simulation.

Then my question is, is that macroscopic monte carlo simulation of silicon detector goes well with real measured data? It will be really helpful if i can get comparison reference or similar things

your answer always help me a lot thanks

6

Hi,
I’m not sure which report you are referring to, but yes, to my knowledge all Garfield++ simulations of LGAD have been done with AvalancheMC so far.
Regarding the agreement with measurements: good question… It’s indeed possible that there are aspects for which one could get more accurate results using a “microscopic” Monte Carlo technique. What application/observable did you have in mind?

8

Sorry it seems that forum recognize my reply as containing some link so i think reply message didn’t transfer to you
Can i ask u what are the points which make microscopic monte carlo not usable in silicon?

9

Sorry, I was a bit unclear. The technique as such is of course valid also for silicon and there are indeed Monte Carlo TCAD device simulation packages on the market. It’s the implementation in AvalancheMicroscopic that is not useable at the moment. Compared to gases, the simulation of the free-flight steps in silicon is more complicated since you need to take the bandstructure into account (and at energies at which impact ionisation is significant the parabolic approximation does not work any more). And of course one needs a reliable set of scattering rates for the different processes.

10

Thanks for your answer! but you say parabolic approximation implemented in garfield is not proper right now. Is that parabolic method based on LD model?

11

Sorry, it’s been a while since I last looked into this. Can you remind me what’s the LD model?
What I meant is that the parabolic approximation is not applicable at high electron (or hole) energies.

12

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.